Predicting atrial fibrillation after cryptogenic stroke via a clinical risk score-a prospective observational study

Area of Science:

• Cardiology
• Neurology
• Medical Diagnostics

Background:

• Atrial fibrillation (AF) is frequently undiagnosed in cryptogenic stroke (CS) patients due to limited cardiac monitoring availability.
• There is a need to pre-select CS patients who would benefit most from extended cardiac rhythm monitoring.
• A clinical risk score was developed to predict AF in CS patients.

Purpose of the Study:

• To develop and evaluate a clinical risk score for predicting atrial fibrillation (AF) in cryptogenic stroke (CS) patients.
• To assess the score's performance over a 1-year follow-up period.
• To improve the diagnostic yield of cardiac monitoring in CS patients.

Main Methods:

• A risk score (0-16 points) was created using variables associated with occult AF, including age, NT-proBNP, ECG/echocardiographic features, and brain imaging markers.
• 150 CS patients admitted between March 2018 and August 2019 were prospectively followed for 1 year for AF detection.
• The score's predictive performance was evaluated using sensitivity, specificity, and negative predictive value.

Main Results:

• During 1-year follow-up, 24 (16%) of 150 CS patients were diagnosed with AF.
• The AF Risk Score (cutoff ≥4) demonstrated a 92% sensitivity and 67% specificity for 1-year AF prediction.
• The score achieved an excellent negative predictive value of 98%, with only two patients scoring <4 diagnosed with AF.

Conclusions:

• A clinical risk score effectively predicts 1-year AF risk in CS patients with high sensitivity and negative predictive value.
• The score aids in selecting CS patients for cardiac rhythm monitoring.
• Further validation in external cohorts with continuous monitoring is necessary to confirm generalizability.